MCS' E-forum debate: collaborative engineering design

Design and manufacturing teams working together over the web is an image in every engineer’s mind today. It’s being touted by the CAD vendors as a key competitive weapon – a way for firms to get more agile, slicker in design and faster to market for new products. It promotes the sharing of design data with other departments, manufacturing engineers, marketing staff, even remote design teams. The concept isn’t new and it certainly isn’t rocket science. As a Manufacturing Management graduate seven years ago, I recall having simultaneous and concurrent engineering rammed down my throat! The difference today is that companies really do have a useful set of tools – the Internet and collaboration software. But who’s using it, and what benefits can users expect? The answers to these questions and others were debated at MCS’ round table forum at Brands Hatch in November last year. The event was run in association with MCS’ sister engineering publication Eureka and the DTI and CBI. Sponsors included: Solidworks, CoCreate, Delcam, Smarteam, Man & Machine (Autodesk reseller), MatrixOne, Mapics, SAP and Oracle. And there were 11 users (heads of design and technical managers), five analysts and consultants. The goal was to uncover the real issues for engineers working collaboratively, what’s behind the vendor acronyms, the kind of internal effort and investment required and how far manufacturing users are prepared to go with this technology. First off, a dose of reality from Guy Washer, director of market research firm Benchmark: “In our latest survey of 1,800 users, 46% said they shared design data, but only 12% said they shared a single, common, live version of a design on-line. And only 14% said they had collaborative design groups involving key customers and/or suppliers.” So UK manufacturing just isn’t doing it on the scale vendors might say. What is collaboration? Nick Ballard, consultant at engineering analyst firm Cambashi kicked off the debate: “Collaborative engineering to me is the way that design information passes through a company’s value chain. At its simplest, even sending drawings on floppy disk to each other is collaboration. It’s getting the right design information to the right people in the right form at the right time.” But Nigel Montgomery, analyst for AMR Research said: “I don’t believe it’s just about information sharing, I think it’s about shared decision making, because you can share design information with people and still screw up. The idea of collaborative engineering is working together to make sure we make smarter decisions. I’d say it’s shared decision making towards a design end.” Most of the table agreed with this. Enter the users. Armando Canales, technical manager at global automotive design company MSXI, said: “We have about 350 CAD workstations across Europe, and we use every system that is out there on the market, because that’s what our clients demand. We’ve evaluated a great number of systems but we’ve not found sufficient acceptance amongst the users to implement them. And they’re not cheap.” So collaborating over the web isn’t easy if you have a multitude of different CAD systems within the business. “We haven’t been able to come across the right cost benefits - this is how much we’re paying for it and this is how much we’re going to get out of it. So we’ve not arrived there yet,” added Canales. Mark Findlay, CAE manager for automotive engineering group Ricardo, said: "My difficulties are that I need to interface with my customers who are all the big automotive firms. My worries are more about getting 2Mb connections in, agreeing security... and often you find the automotive companies aren't geared up for outsourcing thousands of drawings. They're happy to do a single part that's got a couple of drawings you can pass back and forth. But if you're working on an under-hood engine assembly of 1,000 parts, those systems cripple you so you end up doing it on paper. The key question is what do you want from collaborative engineering. For us, it's really the mock-up sessions - that's much more powerful." Another user, James Ewing of Weston Aerospace, explained why his firm had not yet adopted web collaboration software: “At the moment I’m waiting to see how the market pans out. We’re not yet ready to join.” And when asked if he was getting a clear technology message from the vendors, he quipped: “The only clear message I’m getting at the moment is that there’s not something I could easily pick and be sure I’d picked the right thing. It’s not easy to differentiate vendors.” And Graham Hatley, chief designer at aerosol manufacturer Bespak, was another user considering implementing web-based design software: “It’s very easy for me to give data to a supplier, or e-mail it and say that’s collaborative. But if he’s got a problem how does he get it back to me? We just transfer data to our suppliers - we’re trying to find a collaborative solution and are part way through that exercise. We want to pull in different [remote] offices and ultimately spin it out to our major customers and suppliers. We want to put data where they can see it and both look at it and comment on it at the same time. What we don’t want is a time lag working with our tool makers - we send them a file and they say ‘no, we can’t do this’ and they send it back and we change it. We want to be able to speed that process up. So in my mind collaboration is two-way.” But other users are making great strides with the technology. Gus Desbarats, md of UK-based design specialist Total Alloy Product Design, said that true collaboration goes further than simply exchanging design data: “In the consumer electronics industry that we work in there’s actually a huge amount [of design collaboration]. We collaborate with mechanical engineers in China and Taiwan. It’s the decision making and how you go about achieving a decision together in different locations – that’s what it’s about. That’s what is most difficult to implement. But we’re doing it.” Managerial problems too So come on analysts and consultants, what else is holding back adoption? David Tarbox-Cooper, consultant at QIC Solutions, suggested there are two problems at the heart of it. “The first one is we don’t have the technological vehicles to actually implement them properly. The other is purely managerial - simple things like who owns the data, who takes responsibility for certain processes.” Ah, we’re getting somewhere now. Desbarats agreed: “A lot of this can’t be solved by IT alone. A lot of it involves project management techniques and culture. It’s very stressful if you’re a designer to have to comment in real time because whatever you say will have an immediate impact. So when we collaborate with Hewlett Packard on their PDAs [personal digital assistants], it’s a very rapid interactive development process. They’re developing the internal detailing on a frame and we’re developing the pressings around the outside, and the circuitry inside is changing daily. “One of the reasons I think collaborative software isn’t taking off to the extent that people hoped is that there are so many manual work-arounds. CAD allows you to move things along in parallel more efficiently than you have before, because you can communicate more, but there’s still an element of sequence to it.” Mike Askew, general manager at aerospace manufacturer Westland Transmissions, gave a different perspective. “For us, the engineering process is not finished when you get to the geometry. We design sub systems, there’s often a multi-layer structure in the supply chain for design and supply of product for a major assembly. “We’re under very tight timescales to deliver. We’re releasing geometry so [suppliers] can cut dies for castings well before we’ve done the analysis of that part, so it starts to get complicated. What we’d like to do is compress the analysis time for things like stress, thermal analysis and oil flow. So the issue is compressing this analysis time after we’ve got the geometry.” So with all these issues to think about, what is actually driving companies to adopt web technologies for design collaboration? Simon Booker of design software vendor Solidworks: “Ease of use, cost and speed.” He cited an example: “My mother can do Internet shopping quite happily. Why? Because she can use a standard browser: it’s easy to use and it doesn’t necessarily cost her anything, plus it’s fast enough to be usable. The real explosion and use of collaboration tools will come when all those things are addressed.” AMR’s Montgomery: “One [driver] is time to market, that’s what’s driving adoption. People concentrate on the design, but they forget the bit about the supply. So the idea of ‘designing to supply’ is another key factor here - that’s what makes us better than the competition. Instead of just isolating to the best-designed product, you’ve still got to be able to deliver it in the most economical way. So there’s a disconnect between engineering and the real world if you like, and there isn’t a single vendor out there who can solve this.” After a deathly silence from the vendors around the table, Cambashi’s Nick Ballard eventually agreed that the vendors need to make life easier for the user community: “The proliferation of acronyms out there is very unhelpful. It might help the vendors to differentiate between themselves, but it does nothing for the users to actually describe what they can expect from any particular vendor. Everybody has a particular view.” Peter Penhallow, founder of EDM Consultancy, has been working closely with manufacturing firms and had an equally strong message for the vendors: “I liken all these vendor solutions to the cures in a doctor’s bag. Trouble is, who’s analysing the patient [user] and does the patient know what’s actually wrong with him?” Good point indeed but aren’t the users also to blame? “Yes, we’ve got a user community today that doesn’t voice its problems,” said Penhallow. “There’s a gap between what the vendors are saying and what the customer wants. You need a doctor in the middle if you like to diagnose. Getting the users more aware of what their needs are and voicing those more eloquently, rather than the user asking the vendor, ‘what does your product do?’ and ‘Perhaps it will fit our problem’. “IT business is doing itself a disservice by glossing over this issue,” he added. “There’s some fundamental changes to the knowledge, the training of the people, the professional roles, the way you build teams. And unless companies actually face up to these, they won’t get the desired results.” Concurrent benefits So what about concurrent engineering? Desbarats’ example: “In electronics, we discuss with the electronics engineers before they design the board roughly what headroom space they’re going to get, and we break the board up into headrooms and agree an area allocation for each. That’s a frozen requirement, we take that and iterate on that and come up with design options. A connector might change, things might change on that but at least we’re sort of converging on an agreement. “We don’t care what the components are within those spaces, so you freeze assumptions as you work together. And what’s beautiful about 3D – it changes the way you do this – is that you can visualise areas of responsibility. You can create 3D data sets that represent the space you’re allowed to go in like they do on aircraft. You create a cable channel and the big red tube is the thing that the wiring people are allowed to work within, and they only call somebody if they go outside. “It’s nothing to do with the web, it’s the base systems that people are using. The format is the thing and it needs to be a cheap format. This is what’s been discussed for years but still not solved. The important issue with collaboration is that the web is just a medium and there are two ways to shift stuff around, attached to e-mails and data transfers. We use both depending on size. You don’t really need sophisticated IT to do that.” Solidworks’ Simon Booker said: “I think a lot of our customers have found the web a great mechanism for collaborative engineering, not just in CAD but in integrating that with their manufacturing processes, being able to put the design onto the web, and giving either the customers or trading partners access to viewers which give people the ability to look at an object in 3D in a standard Internet explorer browser without an application. You can have one central repository for storing large files that hasn’t got to be communicated around the world to lots of different networks.” Another user, Michael David, director of manufacturing at cash drawer-maker Cash Bases, said: “All our designers are in one location [Newhaven]. One of the huge benefits we’ve found is not purely communicating with engineers, but actually communicating with customers. We manufacture around 2,500 cash drawers each week, average batch size is 10, so you’re looking at 250 orders for different drawer configurations. “You can send a customer a set of drawings, [but] they’re not always engineers, and one of the beauties that you have is being able to send a solid model using an e-drawing. The very fact they can hold it and it’s almost as if it we’re in their hands – it communicates concepts very quickly and removes the need for costly prototypes to be made. These are huge benefits.” ASP the answer? On to IT infrastructure required to run it all. Is this holding users back? Simon Leek, business development manager for CAD/CAM vendor Delcam, suggested that using an ASP (application service provider) could be a cost-effective alternative to doing it all yourself. “Many SMEs that need to collaborate need to have Internet access to talk to other trading partners. The barrier is the actual technology that’s required for them to have web services. They’ve got to have firewalls and the technology in place to support it. ASP is the ideal mechanism to collaborate more effectively.” But when asked how many Delcam customers were actually buying this option, his reply was, “Out of 800 customers, 10 have adopted our ASP service to support collaboration projects.” And this is a fair reflection of the poor current uptake of ASP. SAP consultant Ian Welland explained that business processes were another key criterion for successful web collaboration: “There’s lots of good tools, the web is great and you can look at and share data. But when you confront customers about, ‘What’s your existing process and how do you do it?’ ‘What are you trying to achieve?’ ‘What are you trying to make better, therefore where are you trying to get cost savings, time savings, etc?’, time and time again when it comes down to it and saying, ‘What do you do currently?’, that’s where you get the stumbling blocks.” So what’s the answer? “It’s really about direct sharing, decision support and workflow. Web software gives you a huge tool kit, but at the same time gives you more problems because unless you know what you’re trying to solve, what software do you choose?” Ok, but what about the majority of manufacturing companies, those in the lower tiers with less spending power? EDM Consultancy’s Penhallow had a point to make. “These firms haven’t got their house in order yet in terms of their own data management. Putting stuff on the web is then airing your dirty washing. The key areas for me are who has the authority to send data and who has the authority to receive it, and unless you’ve got that you’ve got a mess.” CoCreate’s Steve Tudberry said his firm had a list of five criteria for evaluating end-users. “Users we sell to have to meet certain criteria before we’ll go into a lot of depth with them. If they’re not looking at their data management and know where the data is, then basically it’s a non-starter. So data management is one, do they have a strategy for it and an infrastructure? Do they have a web or networks of sufficient performance to be able to carry the data around it all? Are they located over several sites or are they on one site? And where it’s fragmented in that site and makes communications across it quite difficult, and so on. “The bottom line is if they don’t have this infrastructure, then really getting involved in collaboration is a non-starter. And if you can’t identify your current processes and how your people interact with them, then applying a digital data system is a waste of time.”